Compare / Gemma 3 4B vs GLM-4.5V

Gemma 3 4B vs GLM-4.5V

Which is cheaper, which is more capable, and which does better at each kind of work, from 13 results on 2 sources that measured both models.

GoogleGemma 3 4B Released 11 Mar 2025 · $0.040 in · $0.050 out
Z.aiGLM-4.5V Released 11 Aug 2025 · $0.60 in · $1.80 out
Gemma 3 4B wins 13 too close to callGLM-4.5V wins 9

Of the 13 results both models have, the better value on each. Results whose reported ranges overlap are too close to call.

Results as of 16 November 2026 (catalogue release 2026-11-16-49bdfa045dee). See also Google vs Z.ai, every model from both labs.

Gemma 3 4B or GLM-4.5V? The short answer

Pick GLM-4.5V if quality matters most: it scores higher (Intelligence Index 7.6 against 4.8). Pick Gemma 3 4B if cost matters more: it costs 21 times less per token.

  • Gemma 3 4B is cheaper: $0.042 against $0.90 per million tokens, 21 times less.
  • GLM-4.5V scores higher on the Artificial Analysis Intelligence Index: 7.6 against 4.8.
  • GLM-4.5V is newer: released 11 Aug 2025, against 11 Mar 2025 for Gemma 3 4B.
  • On third-party benchmarks, on the results both models have: GLM-4.5V does better for reasoning and knowledge and writing.

Head to head

4.8
Intelligence Index
7.6
$0.042
Price per million tokens, blendedlower is better
$0.90
131,072
Context window, tokens
65,536

Every fact side by side

Gemma 3 4BGLM-4.5V
Price per million tokens$0.040 in · $0.050 out$0.60 in · $1.80 out
Blended (3 in : 1 out)$0.042$0.90
Cheapest host, blended$0.042 (Parasail, BF16)$0.90 (Novita, FP8)
Intelligence Index4.87.6
Spring Prompt overall––
Output speed, tokens a second––
Context window131,072 tokens65,536 tokens
Released11 Mar 202511 Aug 2025 (newer)
WeightsOpenOpen
Developer based inthe USChina

Prices are the developer's list price where we have read it, otherwise the typical host on OpenRouter; see every model's price. Intelligence and speed from Artificial Analysis.

Which is better for what

Each model's rank among the models on sale today, for each kind of work, from our own benchmarks and licensed sources. Further right is better; the better rank is in green.

Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls 158 of 159 not measured
Reasoning and knowledgeHard questions across science, maths and general knowledge 181 of 187 144 of 187
WritingWhat people prefer in blind comparisons, and judged writing quality 157 of 169 129 of 169
Sticking to the factsSummarising without inventing things, and factual answers 32 of 139 not measured

Every shared result

Each model's best result on every measure both have, grouped by benchmark, with the setting that produced it. Under each measure, every model that benchmark has measured is a grey tick, best to the right, so you can see where the two sit in the field. The headline measures of each benchmark come first; the rest are under "Show all". Green marks the better value; where a source reports ranges that overlap, neither is marked.

MeasureGemma 3 4BGLM-4.5V
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Overall · TextArena rating, higher is better 255 · 186 of 387 measured 1,303 1,352
Business, management and finance · TextArena rating, higher is better 208 · 199 of 380 measured 1,330 1,342
Creative writing · TextArena rating, higher is better 243 · 197 of 385 measured 1,275 1,310
Artificial Analysis reported by Artificial Analysis · 3 results
Artificial Analysis Intelligence Indexindex score, higher is better 246 · 208 of 249 measured 4.8 7.6thinking reasoning
GPQA Diamond% of questions, higher is better 227 · 161 of 232 measured 29.1% 68.4%thinking reasoning
Humanity's Last Exam% of questions, higher is better 194 · 188 of 250 measured 5.3% 6.3%thinking reasoning
Show all 13 results (7 more)
MeasureGemma 3 4BGLM-4.5V
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Expert prompts · TextArena rating, higher is better 270 · 156 of 338 measured 1,260 1,399
Instruction following · TextArena rating, higher is better 270 · 189 of 387 measured 1,267 1,339
Writing, literature and language · TextArena rating, higher is better 237 · 201 of 386 measured 1,298 1,318
Artificial Analysis reported by Artificial Analysis · 4 results
IFBench% of instructions, higher is better 187 · 162 of 195 measured 28.3% 34.2%thinking reasoning
Long-context reasoning (AA-LCR)% of questions, higher is better 226 · 230 of 241 measured 6.7% 0.0%no reasoning
Terminal-Bench Hard% of tasks, higher is better 180 · 140 of 190 measured 0.8% 6.8%no reasoning
τ²-bench telecom% of tasks, higher is better 187 · 160 of 193 measured 5.0% 22.5%thinking reasoning

Quick answers

Is Gemma 3 4B better than GLM-4.5V?

Pick GLM-4.5V if quality matters most: it scores higher (Intelligence Index 7.6 against 4.8). Pick Gemma 3 4B if cost matters more: it costs 21 times less per token.

Which is cheaper, Gemma 3 4B or GLM-4.5V?

Gemma 3 4B is cheaper: $0.042 against $0.90 per million tokens, 21 times less.

Which is better for reasoning and knowledge, Gemma 3 4B or GLM-4.5V?

GLM-4.5V ranks 144 of 187 models on sale for hard questions across science, maths and general knowledge, against 181 for Gemma 3 4B.

Which is better for writing, Gemma 3 4B or GLM-4.5V?

GLM-4.5V ranks 129 of 169 models on sale for what people prefer in blind comparisons, and judged writing quality, against 157 for Gemma 3 4B.

Which has the bigger context window, Gemma 3 4B or GLM-4.5V?

Gemma 3 4B: 131,072 tokens, against 65,536 for GLM-4.5V.

Run Gemma 3 4B and GLM-4.5V on your own prompt.

Benchmarks aren't your task. Test both side by side on your prompt, with every answer checked against your rules and the cost of each one.